| OCR Text |
Show Journal of Neuro- Ophllmlmohsy I9(. i): 15.1 159. 1999. ( 0 1999 l. ippincoll Williams & Wilkins, Inc., Philadelphia An Anisocoria Produces a Small Relative Afferent Pupillary Defect in the Eye With the Smaller Pupil Byron L. Lam, MD, and H. Stanley Thompson, MD Objectives: To determine whether an anisocoria can produce a relative afferent pupillary defect of clinical importance. Material and Methods: Anisocoria and relative afferent pupillary defect were measured with infrared videography in three clinical experiments: I) every few minutes in eight normal subjects who remained in darkness as one pupil was dilating from mydriatic drops; 2) every 2 hours, for 8 hours in six normal subjects who remained in room light after one pupil was dilated with mydriatic drops; and 3) before and after dilation of one pupil in 24 patients with known afferent defects from optic nerve disease and who remained in room light. Results: In the presence of an anisocoria, the relative afferent pupillary defect was almost always in the eye with the smaller pupil. The results of the three experiments were: I) In darkness, the induced pupillary defect was found to be related to the ratio of the areas of the two pupils ( R = 0.942), and 0.14 log unit of pupillary defect was produced in the eye with the smaller pupil for every millimeter of anisocoria. 2) In room light, the induced pupillary defect was in the eye with the smaller pupil but was less than in Hxpcrimcnl 1 and persisted throughout the 8 hours. This was presumably because the eye with the larger pupil had become more light adapted in the clinic light than the eye with the smaller pupil. 3) In room light, inducing an anisocoria in patients with preexisting afferent pupillary defect tended to shift the pupillary defect toward the eye with the smaller pupil ( R = 0.68). Conclusions: Clinically, approximately 0.1 log unit of relative afferent pupillary defect is produced in the eye with the smaller pupil for every millimeter of anisocoria. Therefore, the anisocoria must be larger than 2 mm in diameter difference to induce a clinically significant relative afferent pupillary defect. Key Words: Anisocoria- Afferent pupillary defect. Whether an anisocoria can, all by itself, produce a significant relative afferent pupillary delect is important in situations in which visual loss and a large anisocoria occur together. For example, you are asked to examine a stuporous patient in the emergency room because his right eye has been injured. The pupil of the injured eye is Manuscript received February 23, 1999; accepted May 26, 1999. From the Basconi Palmer Hye Institute. University of Miami, Miami, Florida ( BLL); and the Department of Ophthalmology, University of Iowa College of Medicine, Iowa City. Iowa ( HST). Address correspondence to Dr. Byron I.. Lam, 900 NW 17th Street, Miami, FL33I36. dilated to 8 mm and unreactive to light, possibly because of a traumatic iridoplegia. Your immediate concern is whether the retina or optic nerve has also been injured. The other eye is uninjured and its 3- mm pupil seems to be reacting normally. You carefully compare the direct and consensual light responses of the left pupil and find them equal. Can you safely accept this as evidence thai there is no traumatic right optic neuropathy? Or could the anisocoria be covering up a real afferent defect? The answer to this question may influence the decision to treat the patient with high- dose corticosteroids or to consider decompression of the optic canal. An anisocoria would be expected to influence pupillary light reactions because the pupillary inequality allows more light to gel into one eye than into the other. The difference in retinal illumination between the two eyes should be proportional to the difference in pupillary area. Thus, the retina of the eye with the larger pupil should receive more stimulus light while the pupils are being tested for a relative afferent pupillary defect. However, in ordinary ambient daylight, the retina of the eye with the larger pupil should become relatively more light adapted and therefore somewhat less sensitive to light than the other eye. It has been clinically estimated that for each millimeter of anisocoria, approximately 0.1 log unit of relative afferent pupillary defect is produced in the eye with the smaller pupil ( 1). This estimate was based on a few measurements of the effect of anisocoria on the Pulfrich illusion and the perception that, given its range of movement, the pupil is unlikely to change retinal illumination by more than 1 log unit. However, careful measurements have not been made in a clinical setting. SUBJECTS AND METHODS Three clinical experiments were performed, and in all instances, an anisocoria was induced by dilating one pupil with eye drops. The size of the pupils and the relative afferent pupillary defects were measured by using an infrared sensitive video system ( 2). The relative afferent pupillary defect was measured by watching the direct and the consensual responses of the undilated pupil that still had a working sphincter. This / w 154 B. L. LAM AND H. S. THOMPSON was performed in a darkened room using a halogen trans-illuminator ( No. 41 100 FinholT; Welch Allyn, Skaneate-lcs Falls, NY) at maximal brightness while the subjects looked at a distant target. Infrared- blocking filters were placed over the stimulus light to avoid washing out the video picture. Calibrated neutral density filters in 0.3 log unit steps were used to measure the afferent pupillary defect. To minimize variation in illumination, one observer alternated the light from eye to eye and kept the light approximately 4 cm from the iris and below the visual axis. At the same lime, the other observer watched the pupillary responses on the screen and decided whether a dense enough filter had been used to reach the balance point. Sufficient filters were used to deliberately overshoot the balance point so thai the true balance point could be determined later from the video tape to estimate the afferent pupillary defect by interpolation to the nearest 0.1 log units. Four to five additional alternations of the light stimulus were then made without the neutral density filters so that the anisocoria could be assessed without the fillers in place. Pupillary size measurements were made by affixing adhesive paper millimeter rules to each lower eyelid in the iris plane. Pupillary diameters were measured from the videotape by using a caliper on the video image of the pupil and reading off the millimeter scale below each eye. The diameters of the pupil in the horizontal and vertical axes were recorded, and the mean of these two diameters was used as the diameter of the pupil. The difference between the pupillary diameters was used as the anisocoria. The left pupil was measured just as the light was leaving the right eye and just before the stimulus light had reached the left pupil and vice versa; that is, in the middle of the 300 milliseconds required to move the stimulus light From one eye to the other, as the light crossed the nose. We did this because the pupil size at the moment that the light fell on the eye determined the retinal illumination and the resulting pupillomotor input. Pupillary diameters were also measured when the neutral density filters needed to balance the pupillary reactions were in place. In Experiment 1, the afferent pupillary defect was measured at different levels of anisocoria during the process of unilateral pharmacologic mydriasis in darkness to determine whether an anisocoria can produce an afferent pupillary defect. Eight adults with normal eyes, equal pupils, and no afferent pupillary defect had their right pupil dilated with a mixture of tropicamidc 0.5% and phenylephrine 5%. The subjects were not dark- adapted before the experiment, and the afferent pupillary defect was measured every few minutes as the right eye dilated. The session was terminated when the medicated eye reached peak mydriasis after 30 to 40 minutes. In Experiment 2, the presence and persistence of an anisocoria- induced afferent pupillary defect in room light was studied. Six normal adults had the right pupil dilated with cyclopentolate 1% and phenylephrine 2.5%. The subjects were dark- adapted for 35 minutes to minimize any preexisting asymmetry of retinal photochemi- ./ Neiiro- Ophllmlmol. Vol. IV. No. J. IW9 cal adaptation. The subjects then returned to work in their usual indoor lighted environment. With the right pupil fixed in mydriasis, the anisocoria and the relative afferent pupillary defect were measured every 2 hours until 8 hours after dilation. Mydriatic drops were placed in the right eye a second time, 4 hours after the first drop to maintain the anisocoria. In Experiment 3, the effect of an induced anisocoria on an existing afferent pupillary defect was studied. The afferent pupillary defect and pupil size were measured in 24 adult patients with preexisting afferent pupillary defects before and after dilating one eye with tropicamide 0.5% and phenylephrine 5%. During the 40 to 60 minutes of dilation, the patients sat in a well- lit waiting room. Any preexisting physiologic anisocoria was subtracted from the final anisocoria. The video segments were reviewed in a random order to determine the amount of the relative afferent pupillary defect for each patient. The results were masked and the videotape was reviewed again to measure the anisocoria. RESULTS In Experiment I, the changes in the relative afferent pupillary defect during mydriasis were measured in darkness. The results of a typical subject are shown in Figure 1. The induced relative afferent pupillary defect was found in the eye with the smaller pupil, and there was a striking linear increase in afferent pupillary defect ( in log units) with increasing anisocoria ( in millimeters of diameter) for all eight subjects. The average number of data points contributed by each subject was 17 ( range 10- 22). The intercepts ranged from - 0.13 to 0.05, the slopes ranged from 0.10 to 0.16, and the correlation coefficients ranged from 0.93 to 0.99. When the results of the eight subjects were combined ( Fig. 2), the pooled estimate of the correlation coefficient ( R) based on growth curve analysis ( 3) was 0.971; with the intercept at zero, the slope was 0.136 ( p < 0.0001), indicating that each millimeter of anisocoria produced approximately 0.14 log units of pupillomotor input defect. When these afferent pupillary defects for all eight subjects were plotted against the anisocoria measured with the balancing neutral filter in place and expressed as the ratio of the area of the larger pupil to the area of smaller pupil ( Fig. 3), the results approached the theoretical expected values ( see Discussion) with the pooled estimate of the correlation coefficient ( R) of 0.942. In Experiment 2, the changes in the relative afferent pupillary defect after mydriasis in room light were assessed. All six normal subjects had a relative afferent pupillary defect of less than or equal to 0.4 log units in the undilated eye after 2 hours in room light ( Fig. 4). This small amount of pupillomotor input asymmetry persisted throughout the experimental period even though the mean afferent pupillary defect showed a slight but statistically significant decrease after hour 5 ( p = 0.0006, analysis of variance). The mean decrease in the afferent pupillary defect throughout the course of experiment was only approximately 0.1 log unit and therefore of doubtful clinical significance. The anisocoria remained relatively ANISOCORIA- INDUCED AFFERENT PUPILLARY DEFECT 155 Q >, as ^ PL, 0 PH - 4-> < L> M- H <! 0) > - t- J ( rt ni Pi ' H P hn o ^ W> , a; cti Ti £ ri 2 3 4 5 6 Anisocoria in Diameter Difference ( mm) FIG. 1. The data of a typical session in Experiment 1. As the anisocoria increased in a normal subject in darkness, so did the induced relative afferent pupillary defect in the undilated eye. constant during the experiment so that the variation of anisocoria was not the cause of the change in the afferent pupillary defect, ( p = 0.57 by analysis of variance with anisocoria as covariable). In Experiment 3, the changes in the relative afferent pupillary defect after mydriasis in patients with existing afferent pupillary defects were examined. The changes in anisocoria in 24 patients were compared to the changes in afferent defect ( Fig. 5). After dilation, the relative afferent pupillary defect showed a tendency to shift slightly toward the undilated eye. The correlation between the amount of change in relative afferent pupillary defect and the amount of change in anisocoria was low, R = 0.68 with linear slope = - 0.074. DISCUSSION When a large anisocoria is found with a relative afferent pupillary defect, it is clinically important to decide how much of the pupillary defect is induced by the anisocoria. Of course, measuring the relative afferent pupillary defect is possible with only one working iris sphincter. The three clinical experiments in this study showed that an anisocoria produces a small afferent pupillary FIG. 2. Results from Experiment 1 with the data points for all eight normal subjects plotted. In darkness, approximately 0.14 log unit of relative afferent pupillary defect is produced in the undilated eye per millimeter of anisocoria. O) 0) p >, ffl ^ a. 3 PH ./) V, 3 hr> o w < < A • n A A R - 0 . 9 7 A A AAAAA A Slope = 0.14 log u n i t s / mm A A H A ^' A AA AA,,"'* A A A A A A ^ A A A A AA AAAA^ lA A A A A AA AAA A A AA , ' 1 ^ 1 A AAA A A A'^ A A A A A A ^" A A A A 1 AA'AAAA A A A A A A 1 2 3 4 5 6 Anisocoria in Diameter Difference ( mm) .1 Neuro- Oplilhalmol. Vol. 19, No. .!. IW9 156 B. L. LAM AND H. S. THOMPSON D != i H. 3 PH ., feren < > 13 o ^-^ 0) >> W lated • n .5 P4 7 , Log ( Area Ratio) R - 0.94 * A- A A 4 8 12 Anisocoria in Area Ratio ( Area of larger pupil / Area of smaller pupil) FIG. 3. Results from Experiment 1. The anisocoria scale has been converted to a ratio of the larger pupil area to the smaller pupil area ( area ratio). Pupillary areas were derived from diameter measurements made with the balancing afferent pupillary defect filter in place. The theoretical curve was calculated from the logarithm of the area ratio. Each 0.3 log unit of neutral density filter cuts the transmission of light in half. The data for the eight subjects approach the theoretically expected values ( see Discussion). defect in the eye with the smaller pupil in normal subjects and patients. The difference in pupillomotor input between the two eyes in a normal subject should be directly related to the difference in the illumination of the two retinas. Retinal illumination is expressed as the intensity of the stimulus light times the pupillary area. This means that assuming a constant intensity of the stimulus light, a constant state of retinal adaptation, and intact afferent pathways, the relative afferent pupillary defect should be related to the ratio of the pupillary areas. Because the relative afferent pupillary defect is measured in log units with neutral density filters, the theoretical relationship between the relative afferent pupillary defect ( RAPD) and the anisocoria would be: RAPD = log Area of larger pupil X stimulus intensity Area of smaller pupil x stimulus intensity log D2 d2 where D = diameter of larger pupil, d = diameter of smaller pupil. However, as the pupil increases in size, the Stiles- Crawford effect and the optical blur of the peripheral lens may affect the pupillomotor input by reducing the effectiveness of the eccentric rays entering the periphery of the pupil ( 4,5). In addition, in ordinary ambient daylight the retina of the eye with the dilated pupil should become relatively light adapted and less sensitive to light. This would tend to reduce the pupillomotor input asymmetry caused by the anisocoria. In Experiment 1, we measured the amount of afferent pupillary defect induced by various amounts of anisocoria during the process of unilateral mydriasis in eight normal subjects. This experiment was performed in darkness without the influence of continuous ambient light. When the anisocoria measured with the neutral density filter in place ( the one that balanced the pupillomotor input asymmetry) over the dilated pupil and expressed as area ratio was plotted against the afferent pupillary defect, the result approached the theoretical values expected from the difference in pupillary areas alone ( Fig. 3). These pupillary measurements with the balancing filter in place represent the anisocoria at the moment the relative afferent pupillary defect was being measured. Of course, these anisocoria measurements with the filters in place were made under special conditions unavailable to the clinician. In the clinic, the pupils are usually measured in millimeters of diameter with a pupil gauge and in diffuse light bright enough to see the pupils. With this in mind, we also measured the anisocoria without any filters in front of the dilated eye. In these subjects, one pupil was dilated and fixed, and the stimulus light was alternated from one eye to the other. The mobile pupil of the undilated eye was measured in a freeze frame just before it had a chance to dilate in response to the removal of the light stimulus from the dilated eye. Therefore, the anisocoria measurements are similar to that which might have been obtained clinically in bright light. When this anisocoria was plotted against the induced relative afferent pupillary defect ( Fig. 2), we found that 1 mm of anisocoria produced approximately 0.14 log units of afferent pupillary defect in the eye with the smaller pupil ( R = 0.971). This constant of 0.14 log units of pupillary defect/ millimeter of anisocoria is influenced somewhat by the range of the pupil size in our subjects because the difference in retinal illumination between the two eyes is related to a difference in pupillary area rather than a difference in diameter, and the same diameter difference for small pupils represents a larger difference in area J Neuro- Ophlhalmol, Vol. 19. No. 3. 1999 ANISOCORIA- INDUCED AFFERENT PUPILLARY DEFECT 157 Relative Afferent Pupillary Defect in Undilated Eye ( log units) Time After Dilation ( hours) Undilated Pupil Diameter ( mm) Mean Anisocoria ( mm) Mean Afferent Defect ( log units) FIG. 4. Results from Experiment 2. The relative afferent pupillary defects in six normal subjects with one dilated eye were measured after exposure to room light. The induced afferent pupillary defect was in the undilated eye and persisted for several hours. ratio for smaller pupils, as compared to larger pupils. Thus, an anisocoria of 2 mm would tend to produce a larger relative afferent pupillary defect for small pupils than for larger pupils. In Experiment 2, the presence and persistence of anisocoria- hiduced afferent pupillary defects in room light were demonstrated in six normal subjects ( Fig. 4). The mean diameters of the undilated and dilated pupils were 3.9 mm and 7.7 mm, respectively, and were relatively constant throughout the 8 hours of the experiment. The induced afferent pupillary defects were 0.4 log units or less in the eye with the smaller pupil and persisted throughout the 8 hours. The amount of afferent pupillary defect was less than the 0.6 log units calculated from the above theoretical formula. In Experiment 3, the effect of a change in anisocoria in room light on 24 clinical patients with known afferent pupillary defects was examined ( Fig. 5). Again, after dilating one pupil, the asymmetry of the pupillomotor input tended to shift toward the undilated eye. Although this shift appeared to be greater in some patients than in other patients and the correlation coefficient was low ( R = 0.68), the result nevertheless suggests that an anisocoria may modify an existing afferent pupillary defect. The slope was - 0.074, which means that approximately 0.08 log units of afferent pupillary defect was produced for every millimeter of anisocoria. This is again less than the 0.14 log units of pupillary defect/ millimeter of anisocoria found in Experiment 1. The anisocoria- induced afferent pupillary defects were smaller in Experiments 2 and 3 than in Experiment 1 and smaller than would be predicted from the theoretical relationship. We presume that this was because, in these two experiments, the subjects were in room light and the retina of the eye with the dilated pupil had become relatively light adapted and less sensitive to light than the undilated eye because more light had been entering the larger pupil. This would tend to reduce the pupillomotor input asymmetry caused by the anisocoria. Nevertheless, the anisocoria- induced relative afferent pupillary defect in all three experiments tended to be in the eye with the smaller pupil, suggesting that the difference in retinal illumination during the alternating light test is the more potent factor and that it more than compensates for any difference in photochemical light adaptation between the two eyes. Experiments 2 and 3 indicate that, under clinical conditions, the afferent pupillary defect shifts < 0.1 log unit ./ Neuro- Ophllwlmol, Vol. 19. No. .1 1999 158 B. L. LAM AND H. S. THOMPSON 1.2 - 1.2' ent SH M- H < tive c3 P< 4u .5 < u tuo X, u C 3 t> JD O) • 4H 0) Q I- I Ctf jzj 3 PH 0) >^ m - M Xi 60 s T j u n3 O H 01 m Toward Left 0.9" 0.6 J 0.3- o. o- - 0.3 ~ - 0.6 " - 0.9- « S> ^ - - ^ "* -'^.. • • ""~" 1s)-~- - / S) « « ,, >. • • ® R = 0.68 Slope = 0.074 log units/ mm A Dilated Good Eye • Dilated Bad Eye O Pre- existing Defect = or < 0.6 log units • ® A ® --. ® ® ® ® ® ® - 1 Left Eye Dilated Right Eye Dilated Change in Anisocoria ( mm) FIG. 5. Results from Experiment 3. The change in anisocoria after dilating one eye was plotted against the change in the relative afferent pupillary defect for 24 patients with a preexisting afferent pupillary defect. The anisocoria tended to shift the relative afferent pupillary defect toward the undilated eye. in the direction of the smaller pupil for every millimeter of anisocoria. By itself, a relative afferent defect of < 0.3 log units carries little clinical weight. Therefore, only large amounts of pupillary inequality (> 2 mm) need to be taken into account when trying to measure the afferent pupillary defect. For example, if one pupil is 9 mm wide and the other is 3 mm, a substantial anisocoria- induced afferent pupillary defect of approximately 0.6 log units should be expected in the eye with the smaller pupil. If, in this situation, no afferent defect is found, the implication is that there is an input defect in the eye with the dilated pupil. The difference between the clinically measured anisocoria and the anisocoria during afferent pupillary defect measurement depends on many factors: whether it is the larger pupil or the smaller pupil that is fixed to light, whether it is the eye with the larger or the smaller pupil that has the input defect, the size of the input asymmetry, and the intensity of the light used while measuring the anisocoria and the relative afferent pupillary defect. In the clinic, the afferent defect is more often found in the eye with the large, poorly reactive pupil caused by injury, third nerve palsy, or previous eye drops. In these cases, the anisocoria varies with lighting conditions. The smaller mobile pupil will constrict as more light enters the pupil, and the brighter the light the greater the anisocoria will be. However, in very bright light, the small pupil will " hit bottom" so that the anisocoria no longer increases with further increases in stimulus intensity ( 6). The act of measuring the afferent pupillary defect may also affect the anisocoria depending on whether the larger or the smaller pupil is fixed. In our experiments, when the balancing filter was in place over the dilated eye, the retinal illumination during the test was decreased by the filter and this reduced the anisocoria because with the smaller pupil mobile and the other pupil fixed in mydriasis, every further decrease of the light stimulus in the dilated eye caused less constriction of the smaller .1 Natro- Oplulutlmal, Vol. 19. N„. J, 1999 ANISOCORIA- INDUCED AFFERENT PUPILLARY DEFECT 159 pupil and this decreased the anisocoria. Hence, the clinically measured anisocoria may be somewhat different from the anisocoria present during the alternating light test depending on the intensity of the stimulus. We would suggest that in the clinic, the pupil size is best estimated with a pupil gauge in moderately bright diffuse room light with the patient looking at a distant object. If in patients with dark irises, more light is needed to see the pupils, then further diffuse light can be shone obliquely on the eyes from below. In conclusion, although there are many factors that may influence the amount of anisocoria- induced relative afferent pupillary defect, the results of this study indicate that an anisocoria produces a relative afferent pupillary defect in the eye with the smaller pupil, and clinically, about a decibel ( 0.1 log unit) of afferent pupillary defect is produced for every millimeter of anisocoria. Therefore, it takes a large anisocoria of > 2 mm in diameter difference to induce a clinically significant relative afferent pupillary defect. REFERENCES 1. Thompson HS. Pupillary signs in the diagnosis of oplic nerve disease. Trans. Ophthalmol Hoc UK 1976; 96: 377- 81. 2. Verdiek RA, Thompson HS. Infrared videography of the eyes. ./ Ophlhal Photo 1991: 13: 19- 21. 3. Graybell FA. ' I'heory and Application of the Linear Model. Boston: Duxbury Press, 1976: 456- 69. 4. Van l. oo JA. Enoch JM. The scolopic Sliles- Crawl'ord effect. Vis Res 1975; 15: 1005- 9. 5. Hnoch JM. . Stimulated response of the retina to light entering different parts of the pupil. ./ Oplic Soc Am 1958: 48: 392- 406. 6. Loewenfeld IH, Newsome DA. Iris mechanics. I: Influence of pupil si/. e on dynamics of pupillary movcnicnl. Am .1 Ophthalmol 1971; 71: 347- 62. J Ncurn- Ophthahnol. Vat. 19. No. .1. 1999 |
